In his letter ^{Reference Lally1} John Lally rightly highlights the ongoing issue of limited information on the important clinical topic of novel psychoactive substances (NPS) – also known as legal highs and head shop drugs – and their use by mental health patients. He refers to his prevalence study in community mental health services, which remains, to our knowledge, the only one of its kind. This knowledge gap chimes with the College Faculty of Addictions Psychiatry report on NPS ^{Reference Bowden-Jones, Fitch, Hilton, Lewis and Ofori-Attah2} pointing out that currently mental health services in the UK have no system-wide method to record psychological harm related to club drugs and NPS.

In an effort to estimate local NPS use prevalence rates in patients presenting to acute mental health services in North Devon, we undertook a small retrospective survey of 100 consecutive acute psychiatric presentations (50 crisis team and 50 in-patient admissions) in January and February 2015. The overall prevalence of NPS use was 8%, a little lower than the 13% described by Lally in his community sample, and it was higher in the in-patient group (12%) than the crisis team group (4%). Based on patients' self-reports, Lally found that in 54% of his community patients the substance taken had an adverse effect on their mental state (mainly psychosis). In our acute setting, the supervising consultant psychiatrists felt that in the majority of cases (n=7/8, 87%) NPS use was clinically relevant to the clinical presentation. ICD-10 diagnoses of patients with acute presentation were also predominantly psychotic (n=5/8, 62.5%).

The locality service covers a large catchment area, with a well-dispersed population of about 150 000 living in an area of 420 square miles. Of the seven people living locally, six had residential addresses within a mile of a shop known to be openly selling legal highs; the remaining lived within 2 miles of the shop. There were no people from towns without known legal high shops. This is of potential interest and relevance to any public health or local government interventions.

This was a small sample, with much more simplistic methodology than Lally's study, making any firm conclusions difficult. Given its retrospective nature, and reliance on individual's disclosure and clinician's documentation, our results are likely to be an underestimate of the true prevalence. However, we are aware of no other published record of NPS use prevalence rates in an acute psychiatric population.